Fuzzy match tree labels

ms609 · Jun 26, 2023 · e592d83 · e592d83
1 parent 6d893ba
commit e592d83
Show file tree

Hide file tree

Showing 3 changed files with 11 additions and 3 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -47,7 +47,7 @@ Imports:
   shiny (>= 1.6.0),
   shinyjs,
   stats,
-  TreeDist (>= 2.5.0),
+  TreeDist (>= 2.6.1.9000),
   TreeTools (>= 1.9.2.9002),
 Suggests:
   knitr,

diff --git a/NEWS.md b/NEWS.md
@@ -2,6 +2,8 @@
 
 - Default to use equal weighting during ratchet iterations
 - Support null constraints in `AdditionTree()`
+- Allow search to continue when loading a new file with different taxon names
+  into the app
 
 # TreeSearch 1.3.2 (2023-04-27)
 

diff --git a/inst/Parsimony/app.R b/inst/Parsimony/app.R
@@ -1570,9 +1570,15 @@ server <- function(input, output, session) {
                          " # First tree with optimal score"))
           r$trees[[firstOptimal]]
         } else {
-          # TODO Could try fuzzy matching
+          # Fuzzy-match labels
+          oldTree <- r$trees[[1]]
+          oldLabels <- TipLabels(oldTree)
+          newLabels <- TipLabels(r$dataset)
+          matching <- TreeDist::LAPJV(adist(oldLabels, newLabels))$matching
+          scaffold <- KeepTip(oldTree, !is.na(matching))
+          scaffold[["tip.label"]] <- newLabels[matching[!is.na(matching)]]
           AdditionTree(r$dataset, concavity = concavity(),
-                       constraint = KeepTip(r$trees[[1]], TipLabels(r$dataset)))
+                       constraint = scaffold)
         }
       }
       LogMsg("StartSearch()")